基于K-均距异常因子的步态信号异常检测
|
摘要
针对在提取步态特征时,步态信号的有效采样距离短、模式异常多、难以满足周期划分需求的问题,提出了一种基于K-均距异常因子的步态序列异常检测方法.首先,对步态信号进行自适应小波去噪,通过边缘权重因子提取边缘点划分子模式,然后以4个特征值构建四维特征空间和特征子空间来计算异常因子,最后以异常值均值为标准,以步态周期为单位,对步态序列进行筛选.经公开数据集和自采数据集实验,结果证明在步态信号中检测步态周期模式异常的准确性、合理性和有效性.
To solve the difficulties of feature extraction for acceleration signals,due to the limited number of sampling points and many outlier patterns of gait signals,which can't be used to divide the gait cycle,we propose a novel outlier detection algorithm based on K-mean distance outlier factor detection. We use a wavelet denoising method to adaptively pretreat gait signals,and extract the edge point of the gait series according to the edge weight. This achieves a four-dimensional feature space and feature subspace by the use of the four eigenvalues of each sub-pattern. We then compute the K-mean distance outlier factor. Lastly,we screen the gait series with the mean outlier value,and test it using the unit of the gait cycle. The promising experimental results demonstrate that the proposed algorithm is accurate and reasonable on open access datasets as well as our built datasets,and efficiently detects outliers in gait series.
To solve the difficulties of feature extraction for acceleration signals,due to the limited number of sampling points and many outlier patterns of gait signals,which can't be used to divide the gait cycle,we propose a novel outlier detection algorithm based on K-mean distance outlier factor detection. We use a wavelet denoising method to adaptively pretreat gait signals,and extract the edge point of the gait series according to the edge weight. This achieves a four-dimensional feature space and feature subspace by the use of the four eigenvalues of each sub-pattern. We then compute the K-mean distance outlier factor. Lastly,we screen the gait series with the mean outlier value,and test it using the unit of the gait cycle. The promising experimental results demonstrate that the proposed algorithm is accurate and reasonable on open access datasets as well as our built datasets,and efficiently detects outliers in gait series.
引文
[1] Bchlin M,Plotnik M,Roggen D,et al. Wearable assistant for Parkinson's disease patients with the freezing of gait symptom[J]. IEEE Trans-actions on Information Technology in Biomedicine,2010,14(2):436-446.
[2]朱明.三轴加速度传感器系统在运动技术分析领域的研究与应用[J].冰雪运动,2015,37(2):89-96.Zhu M. Research and application of triaxial acceleration transducer system in the field of sport technical analysis[J]. China Winter Sports,2015,37(2):89-96.
[3]路永乐,张欣,龚爽,等.基于MEMS惯性传感器的人体多运动模式识别[J].中国惯性技术学报,2016,24(5):589-594.Lu Y L,Zhang X,Gong S,et al. Recognition of multiple human motion patterns based on MEMS inertial sensors[J]. Journal of Chinese Iner-tial Technology,2016,24(5):589-594.
[4] Ren Y Z,Chen Y Y,Chuah M C. User verification leveraging gait recognition for smartphone enabled mobile healthcare systems[J]. IEEETransactions on Mobile Computing,2015,14(9):1961-1974.
[5] Khandelwal S,Wickstrm N. Gait event detection in real-world environment for long-term applications:incorporating domain knowledge intotime-frequency analysis[J]. IEEE Transactions on Neural System and Rehabilitation Engineering,2016,24(12):1363-1372.
[6]谷丽华,崔畅,高松巍,等.基于MPU-6050的步态信号采集系统[J].沈阳工业大学学报,2015,37(2):76-182.Gu L H,Cui C,Gao S W,et al. Gait signal acquisition system based on MPU-6050[J]. Journal of Shenyang University of Technology,2015,37(2):76-182.
[7]侯向锋,刘蓉,周兆丰.加速度传感器MMA7260在步态特征提取中的应用[J].传感技术学报,2007,20(3):507-511.Hou X F,Liu R,Zhou Z F. Application of accelerometer MMA7260 in the gait feature extraction[J]. Chinese Journal of Sensors and Actua-tors,2007,20(3):507-511.
[8]章政,王龙,曾聪.运动加速度抑制的动态步长梯度下降姿态解算算法[J].信息与控制,2017,46(2):136-143.Zhang Z,Wang L,Zeng C. Attitude estimation algorithm of motion acceleration restrained with dynamic step gradient descent[J]. Informationand Control,2017,46(2):136-143.
[9] Soaz C,Diepold K. Step detection and parameterization for gait assessment using a single waist-worn accelerometer[J]. IEEE Transactions onBiomedical Engineering,2016,63(5):933-942.
[10]Trung N T,Makihara Y,Nagahara H,et al. Performance evaluation of gait recognition using the largest inertial sensor-based gait database[C]//5th IAPR International Conference on IEEE Biometrics Compendium. Piscataway,NJ,USA:IEEE,2012:360-366.
[11]涂斌斌,谷丽华,揣荣岩,等.基于MEMS加速度传感器的步态识别[J].中国惯性技术学报,2017,25(3):304-308,359.Tu B B,Gu L H,Chuai R Y,et al. Gait recognition based on MEMS acceleration sensor[J]. Journal of Chinese Inertial Technology,2017,25(3):304-308,359.
[12]Wang Y X,Wu K S. Wi Fall:Device-free fall detection by wireless networks[J]. IEEE Transactions on Mobile Computing,2017,16(2):581-594.
[13]Gu Y,Ren F J. PAWS:Passive human activity recognition based on Wi-Fi ambient signals[J]. IEEE Internet of Things Journal,2016,3(5):796-805.
[14]Yu M. An online one class suport vector machine-based person-specific fall detection system for monitoring an elderly individual in a room envi-ronment[J]. IEEE Journal of Biomedical and Health Informatics,2013,17(6):12-14.
[15]Zhang Y T,Pan G,Jia K,et al. Accelerometer-based gait recognition by sparse representation of signature points with clusters[J]. IEEETransactions on Cybernetics,2015,45(9):1864-1875.
[16]李娜,侯义斌,黄樟钦.基于人体加速度特征的实时跌倒识别算法[J].小型微型计算机系统,2012,33(11):2410-2413.Li N,Hou Y B,Huang Z Q. Implementation of a real-time fall detection algorithm based on body's acceleration[J]. Journal of Chinese Com-puter Systems,2012,33(11):2410-2413.
[17]潘纲,张宇霆,吴朝晖.一种基于加速度传感器的步态身份识别方法:中国:200910153244.2[P]. 2009-10-29.Pan G,Zhang Y T,Wu C H. Gait recognition method based on gait acceleration:China:200910153244.2[P]. 2009-10-29.
[18]Li Y X,Wang X B,Qiao F. Gait authentication based on acceleration signals of ankle[J]. Chinese Journal of Electronics,2011,20(3):447-451.
[19]陶珂,朱建军.小波去噪质量评价方法的对比研究[J].大地测量与地球动力学,2012,32(2):128-133.Tao K,Zhu J J. A comparative study on validity assessment of wavelet de-noising[J]. Journal of Geodesy and Geodynamics,2012,32(2):128-133.
[20]涂斌斌,谷丽华,许会.步态识别的小波去噪质量评价方法[J].沈阳工业大学学报,2017,39(1):61-66.Tu B B,Gu L H,Xu H. Quality evaluation method for wavelet de-noising in gait recognition[J]. Journal of Shenyang University of Technolo-gy,2017,39(1):61-66.
[21]詹艳艳,徐荣聪.时间序列异常模式的K-均距异常因子检测[J].计算机工程与应用,2009,45(9):141-145.Zhan Y Y,Xu R C. K-mean distance outlier factor detect for outlier pattern of time series[J]. Computer Engineering and Applications,2009,45(9):141-145.
[22]Derawi M,Bours P,Holien K. Improved cycle detection for accelerometer based gait authentication[C]//6th International Conference on In-telligent Information Hiding and Multimedia Signal Processing. Piscataway,NJ,USA:IEEE,2010:312-317.
[23]Pan G,Zhang Y,Wu Z. Accelerometer-based gait recognition via voting by signature points[J]. Electronics Letters,2009,45(22):1116-1118.
[24]Gafurov D,Snekkenes E,Bours P. Gait authentication and identification using wearable accelerometer sensor[C]//Conference on AutomaticIdentification Advanced Technologies. Piscataway,NJ,USA:IEEE,2007:220-225.
[25]胡彬,邵叶秦.基于核学习和距离相似度量的行人再识别[J].信息与控制,2017,46(5):525-529,542.Hu B,Shao Y Q. Person re-identification using kernel learning and similarity-distance metric[J]. Information and Control,2017,46(5):525-529,542.
[2]朱明.三轴加速度传感器系统在运动技术分析领域的研究与应用[J].冰雪运动,2015,37(2):89-96.Zhu M. Research and application of triaxial acceleration transducer system in the field of sport technical analysis[J]. China Winter Sports,2015,37(2):89-96.
[3]路永乐,张欣,龚爽,等.基于MEMS惯性传感器的人体多运动模式识别[J].中国惯性技术学报,2016,24(5):589-594.Lu Y L,Zhang X,Gong S,et al. Recognition of multiple human motion patterns based on MEMS inertial sensors[J]. Journal of Chinese Iner-tial Technology,2016,24(5):589-594.
[4] Ren Y Z,Chen Y Y,Chuah M C. User verification leveraging gait recognition for smartphone enabled mobile healthcare systems[J]. IEEETransactions on Mobile Computing,2015,14(9):1961-1974.
[5] Khandelwal S,Wickstrm N. Gait event detection in real-world environment for long-term applications:incorporating domain knowledge intotime-frequency analysis[J]. IEEE Transactions on Neural System and Rehabilitation Engineering,2016,24(12):1363-1372.
[6]谷丽华,崔畅,高松巍,等.基于MPU-6050的步态信号采集系统[J].沈阳工业大学学报,2015,37(2):76-182.Gu L H,Cui C,Gao S W,et al. Gait signal acquisition system based on MPU-6050[J]. Journal of Shenyang University of Technology,2015,37(2):76-182.
[7]侯向锋,刘蓉,周兆丰.加速度传感器MMA7260在步态特征提取中的应用[J].传感技术学报,2007,20(3):507-511.Hou X F,Liu R,Zhou Z F. Application of accelerometer MMA7260 in the gait feature extraction[J]. Chinese Journal of Sensors and Actua-tors,2007,20(3):507-511.
[8]章政,王龙,曾聪.运动加速度抑制的动态步长梯度下降姿态解算算法[J].信息与控制,2017,46(2):136-143.Zhang Z,Wang L,Zeng C. Attitude estimation algorithm of motion acceleration restrained with dynamic step gradient descent[J]. Informationand Control,2017,46(2):136-143.
[9] Soaz C,Diepold K. Step detection and parameterization for gait assessment using a single waist-worn accelerometer[J]. IEEE Transactions onBiomedical Engineering,2016,63(5):933-942.
[10]Trung N T,Makihara Y,Nagahara H,et al. Performance evaluation of gait recognition using the largest inertial sensor-based gait database[C]//5th IAPR International Conference on IEEE Biometrics Compendium. Piscataway,NJ,USA:IEEE,2012:360-366.
[11]涂斌斌,谷丽华,揣荣岩,等.基于MEMS加速度传感器的步态识别[J].中国惯性技术学报,2017,25(3):304-308,359.Tu B B,Gu L H,Chuai R Y,et al. Gait recognition based on MEMS acceleration sensor[J]. Journal of Chinese Inertial Technology,2017,25(3):304-308,359.
[12]Wang Y X,Wu K S. Wi Fall:Device-free fall detection by wireless networks[J]. IEEE Transactions on Mobile Computing,2017,16(2):581-594.
[13]Gu Y,Ren F J. PAWS:Passive human activity recognition based on Wi-Fi ambient signals[J]. IEEE Internet of Things Journal,2016,3(5):796-805.
[14]Yu M. An online one class suport vector machine-based person-specific fall detection system for monitoring an elderly individual in a room envi-ronment[J]. IEEE Journal of Biomedical and Health Informatics,2013,17(6):12-14.
[15]Zhang Y T,Pan G,Jia K,et al. Accelerometer-based gait recognition by sparse representation of signature points with clusters[J]. IEEETransactions on Cybernetics,2015,45(9):1864-1875.
[16]李娜,侯义斌,黄樟钦.基于人体加速度特征的实时跌倒识别算法[J].小型微型计算机系统,2012,33(11):2410-2413.Li N,Hou Y B,Huang Z Q. Implementation of a real-time fall detection algorithm based on body's acceleration[J]. Journal of Chinese Com-puter Systems,2012,33(11):2410-2413.
[17]潘纲,张宇霆,吴朝晖.一种基于加速度传感器的步态身份识别方法:中国:200910153244.2[P]. 2009-10-29.Pan G,Zhang Y T,Wu C H. Gait recognition method based on gait acceleration:China:200910153244.2[P]. 2009-10-29.
[18]Li Y X,Wang X B,Qiao F. Gait authentication based on acceleration signals of ankle[J]. Chinese Journal of Electronics,2011,20(3):447-451.
[19]陶珂,朱建军.小波去噪质量评价方法的对比研究[J].大地测量与地球动力学,2012,32(2):128-133.Tao K,Zhu J J. A comparative study on validity assessment of wavelet de-noising[J]. Journal of Geodesy and Geodynamics,2012,32(2):128-133.
[20]涂斌斌,谷丽华,许会.步态识别的小波去噪质量评价方法[J].沈阳工业大学学报,2017,39(1):61-66.Tu B B,Gu L H,Xu H. Quality evaluation method for wavelet de-noising in gait recognition[J]. Journal of Shenyang University of Technolo-gy,2017,39(1):61-66.
[21]詹艳艳,徐荣聪.时间序列异常模式的K-均距异常因子检测[J].计算机工程与应用,2009,45(9):141-145.Zhan Y Y,Xu R C. K-mean distance outlier factor detect for outlier pattern of time series[J]. Computer Engineering and Applications,2009,45(9):141-145.
[22]Derawi M,Bours P,Holien K. Improved cycle detection for accelerometer based gait authentication[C]//6th International Conference on In-telligent Information Hiding and Multimedia Signal Processing. Piscataway,NJ,USA:IEEE,2010:312-317.
[23]Pan G,Zhang Y,Wu Z. Accelerometer-based gait recognition via voting by signature points[J]. Electronics Letters,2009,45(22):1116-1118.
[24]Gafurov D,Snekkenes E,Bours P. Gait authentication and identification using wearable accelerometer sensor[C]//Conference on AutomaticIdentification Advanced Technologies. Piscataway,NJ,USA:IEEE,2007:220-225.
[25]胡彬,邵叶秦.基于核学习和距离相似度量的行人再识别[J].信息与控制,2017,46(5):525-529,542.Hu B,Shao Y Q. Person re-identification using kernel learning and similarity-distance metric[J]. Information and Control,2017,46(5):525-529,542.